Bottom Flow: Diffusers and underbodies.. feeding vs starving

[Jersey Tom]

Trying to wrap my head around what I view as two very different ideologies in underbody design.

Internal and external aerodynamics has been fundamentally one of my weak points, never took any classes regarding it. I do know a little. Let me run my understanding of this stuff by, correct me where I'm wrong.

In F1 the underbody and rear "diffuser" create heaps of downforce. As I understand this is mostly because you're effectively creating an airfoil out of the bottom, particularly if the car pitches forward on the brakes. Having the diffuser channels shaped the way they are promotes vortex generation.. allowing the diffuser to run at a real high diverging angle without stalling (breaking up / energizing boundary layer?). Also some decrease in drag by returning the underbody flow to somewhere closer to freestream (though I'd think the big upswirling vortices would negate this.. as opposed to smooth laminar flow?)

Even with most of the bottom of the car being flat I'd imagine you get most of the downforce from a "venturi" effect under the car and upping the velocity to decrease pressure. Not the best venturi since the sides can't be sealed by regs, but not bad nonetheless.

Here is where I get tripped up.. how do you feed that underbody and venturi effectively? I'd think you'd want a "backwards diffuser" toward the front of the car and on the leading edge of the sidepods to ram air under there, and up the air velocity. Ie if you can't feed it, what good is it? But I dont know mucha bout how those forward edges are shaped.

Take a polar opposite though - Nascar. Laugh as you may, aero is a major setup concern in the series, which is why on some tracks they ride on bumpstops the whole damn time. The front splitter is a big deal, and getting the ride height as absolutely low as possible to build up all that stagnation pressure on the top of the splitter, and stop any air from getting under the car.. starving it for air rather than force feeding it.

Why would you do this? I tried looking up some Nascar regs but couldn't find much. I assume they are not allowed to run a flat, smooth underbody and has to be all exposed crap. If that were the case I could see that lots of airflow under the car would just become turbulent rapidly and increase pressure.. raising drag and lift.

So is it just a polar opposite design choice with no middle ground? If you can run a flat underbody and diffuser, get the air down there.. otherwise choke it completely?

[Carlos]

This PDF on Vortex Behaviour in F1 Underbody Conditions might have some hints
http://www.aero.lr.tudelft.nl/education ... 4_1_05.pdf

[Jersey Tom]

Eh. For what it is, that's a decent read.. but it's not what I'm lookin for.

[bettonracing]

You should check out Competition Car Aerodynamics by Simon McBeath. It has CFD data on both situations that should help answer some questions. (I know You keep hearing about this book and although I don't feel like I learned much, it did reinforce the fundamentals and answered a few doubts

The closer a (motorsports) wing is to the ground, usually the more downforce it creates - up to a point of course. Using this logic, the venturi effect will create lower pressure than the choked scenario. You're correct about the Nascar underbody regulations. If it was legal, they'd be doing smooth underbodies. As You speculated, underbody airflow would create higher pressure under the car and thus it's better in their case to choke the flow.

Since most aero cars need a good deal of front downforce, they run the front splitters/bumpers/wings close to the ground, as opposed to a large opening to ingest lots of air for the tunnels/venturi (basically two venturi sections - front and middle). At the same time, You will also notice most aero cars' front splitters/bumpers/wings have a raised center section to help feed the underbody.

In summary: Feed the tunnels. Don't feed the crap. Don't forget to keep the front wing/ splitter close to the ground (which in turn somewhat chokes the tunnels).

I'll stop here and let the experts chime in.

Regards,

Kurt

P.S. A practice becoming more commonplace is to seal underbodies with vortices at the sides. Barge boards & front wing end plates are the two most common places (I speculate) that are used to create these vortices.

[Jersey Tom]

P.S. A practice becoming more commonplace is to seal underbodies with vortices at the sides. Barge boards & front wing end plates are the two most common places (I speculate) that are used to create these vortices.

If that's true.. that's one of the coolest things I've heard. Good input.

I'd like to hear Ogami's take on everything..

[effuno]

P.S. A practice becoming more commonplace is to seal underbodies with vortices at the sides. Barge boards & front wing end plates are the two most common places (I speculate) that are used to create these vortices.

I recently tried a study on inducing vortices on the sides to avoid the high pressure air from forming on the sides and spilling into the diffuser. Frankly, I did not achieve the desired results, but mainly because I couldn't devote enough time and energy towards this to make sure that I do get the results. But, I thought this must have been used by the teams for long time, say, atleast since the bargeboards took the shape they are in now !

Having gone through the process, I did achieve lower pressure on the lower parts of the sides. But this again prompted spillage from the top. Next stage in the study would have been(yes..that never happened ) using that spillage to generate vortices as these vortices-some sort of edge vortices by putting some'end plate' sort of strucuters- could be produced at a point further behind than the point where the side vortices are produced right now : bargeboards.

It wasnt a full car simulation. I left an aerofoil shaped wall curved similarly to the sides of an F1 car. Used an extruded and fillet-ed cicle(simulating wheel) to create wake, simulating the front wheel. Did not model a genuine sidepod 'hole'(thus, skipping the hot internal flows and in turn leaving out the possibility of using any exhaust gases to lower the pressure anywhere else). The only thing that might come close to reality is BARGEBOARD, but even this was done with dimensions found by scaling pictures, and the edges of the bargeboards were created using splines. And, add the fact that first order discretization was used and y plus values nowwhere near unity, this was a simulation that hold no significance with regard to finding exact values. More like helped me find if my thoughts were in right direction, cause I hadn't seen any discussion or article on the side vortices. (if I had, I wudnt have done a simulation..lol)

To conclude (if at all I can use that word), side vortices did give lower pressure, better negative lift in diffusers, higher drag, and pretty much same efficiency(was a bit lower than orig, in fact) for what I did. But again, as I said,modelling sidepods and internal flows, using exhaust gases and creating vortices from the spillage from the top part might add to the lift without much extra drag.

[scarbs]

I think you’ll find with Nascar and other series where wide flat front splitters are in use that the rest of the cars underbody is less than flat or at least at a higher ride height than the splitter. Thus the splitter can be very low without the fear of stalling the entire underbody flow. The exception is the Sports prototypes categories where splitters now need to be stepped to prevent the unexpected stalling of the floor.

Feeding the floor is better, a Diffuser works by expansion, the more flow you can pass through the more downforce you produce, getting flow through the diffuser is dictated by exit area, onset flow, ride height and sealing, plus due to the rules what sits between the tunnel and the inlet (i.e flat floor).

F1 floors are stymied by regulations and their design owes little to the ideal shape for underbody design (see below). Even recent champcars are restricted by the rules, ever smaller tunnels and the upswept sidepod front acting more as location for turning vanes than feeding the diffuser.

The early eighties wing cars were a better shape for underbodies, with a converging inlet feeding a long tall diverging exit, sealed at the sides by skirts.

F1 floor tech
F1 cars are now regulated to limit side skirts, the flatness of the floor and the height of the diffuser exits, this fixes their potential to create downforce.

Sideskirts
F1 have been banned from side skirts since 1983, thus the floor cannot be sealed from the sides allowing air to spill from the flanks of the sidepods under the floor. The coke bottle shape was a step towards creating a low pressure region above the floor and a ‘footplate’ to limit pressure migration. Later teams adopted bargeboards one use of which was to create a vortice to help seal the floor. More recently the undercut sidepods and axe heads extend the foot plate effect of the coke bottle shape to the very front of the sidepod. This latter development helps fool the floor into thinking its wider than it really is, as the complex flow off the front asplitter around the sidepods tends to naturally pass back into the coke bottle shape, with gurney flaps and flip up to keep the flow from passing under the floor.

Ride height
Ride height for the diffuser is almost fixed, the side channels sit 5cm above the middle channel and that itself sits above the plank. The remaining ride height provided by suspension movement (25mm front 75mm rear) is less influential, but keeping the cars rake and ride height still need to be controlled by the third damper. As you say rake allows for increased DF at lower speed before the aero compresses the rear end to flatten out the car at speed to reduce the rear wings angle of attack.

Exits
As the side channels are now limited to a height of 125mm (curiously the centre tunnel is almost unlimited and can be as high as the rear beam wing). So the diffuser is exit limited. Thus teams need to work the exit harder to create more potential from the exits. Thus teams fit aggressive gurneys and work the rear lower wishbones, toe arms and beam wing in cascade with the diffuser. Plus we see teams with more exit area from ever taller centre tunnels and the small supplementary side channels.

Flatness
As the floor cannot have a converging inlet to feed the diffuser, the only way to feed the floor with more mass flow is to use the bargeboards to send high pressure air to the floor in an even manner, thus the appearance of serations and curls on the boards footplate aid pressure under the floor by sending vortices into the floor onset flow. Equally the front face of the sidepods are bluff and near vertical and square to the oncoming airflow. This creates a dam effect and the floor is actually drawing air from several centimetres above its ride height.

[Jersey Tom]

Good stuff, Scarbs. Particularly the last bit on feeding the floor.

Controlling the pitch with 3rd damper is slick, I like it. Soft front compression and rear rebound.. stiff front rebound and rear compression.

Though that might make the front jack down a bit.

[bettonracing]

Scarbs (or any other experts that wish to chime in [Ogami]), do You know if teams are using small diameter vortices (such as those produced by a sharp edge) to simulate the skirt or if they're using large diameter vortices (let's say 150mm diamater or larger, such as those produced by the rear wing) to energize the airflow in an outward direction?

Intuition (looking at the barge boards and the front wing end plates) tells me they're using the smaller diameter, at least smaller than 75mm diameter, to reduce/ eliminate spillage. The larger diameter vortices seem like they would induce a significant amount of drag (and possibly influence the underbody airflow). All that being said, reality and intuition don't always agree...

Regards,

Kurt

[dnomdec]

the only way to feed the floor with more mass flow is to use the bargeboards to send high pressure air to the floor in an even manner, thus the appearance of serations and curls on the boards footplate aid pressure under the floor by sending vortices into the floor onset flow. Equally the front face of the sidepods are bluff and near vertical and square to the oncoming airflow. This creates a dam effect and the floor is actually drawing air from several centimetres above its ride height.

Scarbs, mind explaining more why sending vortices into the floor would send high pressure air to the floor in an even manner? I must be missing something, coz vortices seem to introduce unevenness (don't know if it makes self-steepening flow)
and low pressure...

Ned

[RH1300S]

A question: - reading these posts it seems that people are assuming that the majority of the flow feeding the underfloor is coming from the front.

Look at the diffuser shapes of an F1 car while it is upside down and they suggest a flow coming from the sides at an angle (approx. 45deg) and converging on the diffusers.

My imagination sees some of the air arriving at the heavily scuplted and undercut lower sidepod - creating a little bit of usefull high pressure there, and some of it rolling smoothly off the side and being drawn under the car by the suction of the underfloor.

Whilst (obviously) much flow must arrive from the front; surely accepting that the sides are not sealed and employing the diffusers to pull that flow through seems logical.

This also suggests that the affect of the diffuser being worked by the rear wing (hope I understood that relationship properly ) is actually very powerfull.

Can anyone shed some light on this?

[scarbs]

I can understand your point about flow entering from the side. I guess its inevitable as there’s no side skirts, but this would be air with lower energy and would likely upset the efficiency of the diffuser.
You mentioned the angled design of the tunnels, this is not to make use of flow from the side of the car, but to direct the side channels flow into the taller central tunnel. As the central tunnel is taller it can help pull air from the lower side channels, so this angled design is just to make the side channel area effectively larger

[RH1300S]

Thanks Scarbs - having thought about your answer for a couple of days ( ) - I think I am getting the picture.

It does seem to make sense and in that context my suggestion is wide of the mark.

Back nearer topic

Does that mean that feeding is better as long as you can get the stuff out again via. an effective diffuser? Starving is about minimising the downside of air passing under the car and having nothing better to do than mess things up?